I'd say no.  They're both dependent on the source and receiver characteristics, although they are both possibly damped, which tends to favor lower frequencies.  In a lossless medium, neither body nor Rayleigh waves are dispersive.  Hope that helps.

Surely they both do!

The dominant frequency is mainly dependent on the seismic source and its interaction with the material, lets say rock. Due to attenuation and dispersion you will then measure also different frequency content and amplitudes over time and distance from source, taking for granted that you have an adequate receiver system.

The receiver system has to be able to record the apparent frequencies and amplitude levels. For example if you set up a surface wave array and you want to measure the natural microseisms at their peak, you would probably chose broadband seismometers to detect weak frequencies around 0.1Hz. If you use an active system you will produce higher frequencies and you would for example use airguns and ocean-bottom geophones around 10Hz resonance frequency to record from 3Hz and up to several 100Hz.

Again, all sources (natural or human) have a specific frequency characteristic and the waves (body or surface) will always depend on it. 

They're both dependent on the source and receivers ,

actually you have introduce the frequency range for the isolated source but you dont say anything about the wavefield.

for example asten (1984) said the Oceanic waves striking along the coasts is about 0.5 to1.2 Hz

but the wavefield ( body or surface ) need to more investigation. ( like the SPAC or F-K methods )

Concerning the train as a source, I would suggest to look at frequencies above 1Hz. That limits your depth penetration usingsurface waves, which is probably also dependent on train size and speed. Standard geophones might be just OK, but simpler seismometer will probably give you better data.

The answers are taking several forms, partly because the original question is a little ambiguous.  Let me try a more specific answer.  Body waves and Rayleigh waves propagating in a homogeneous, isotropic, non-dissipative half-space, propagate in a frequency independent manner.  There is no dispersion (phase velocity dependent on frequency) for either body or Rayleigh waves.

If we add a source and receiver to this system, and examine the output of the receiver, the variation with frequency that we see in the resultant signal is due entirely to the frequency characteristics of the source and the receiver.

In real life the situation is complicated by the presence of internal structure in the half-space (such as layering) and of damping in the half-space material.  Both of these produce frequency-dependent effects, in addition to the ones already caused by the details of the source and receiver.

Of course are both wave types dependant on the source and the receiver region characteristics. On top of that is one important criteria the distance of the observational point from the source. Seismic waves have a Q-factor and the higher frequencies are "fatser" damped than the lower frequencies. That means the further away from the source you are, the lower are the observed frequencies and their range. Another important point is the source itself. Explosions have very different frequnecy characteristics rathr than earthquakes. All that info you might find in a good textbook on seismic source and wave propagation.

for all answer thanks a lot

to clarifying let say i meant :  in the same condition of soil , receiver , source . for example in this case  we have a band of frequency  from 1 to 200 Hz and  2 or 3 dominant frequency that is induced from source . i want to know which part of this band is related to body wave or  Rayleigh wave . or  can i distinct  body wave from Rayleigh wave from the contours.